Gather datasets

Data inputs

The cumulative index score includes the following equally weighted inputs.

• Poverty
• Less than high school education
• Linguistic isolation
• Unemployment rate
• Housing burden

Combining data from different geographic units

Some data is not available at the census block group level and is instead only available for larger units such as census tracts or counties. In these cases, all census block groups will get an even contribution from the larger unit. For example, if a census tract scores 90th percentile on an indicator, then all census block groups within that tract will receive a value of 90th percentile.

Normalizing data

The range of the data that makes up the score varies, so the data must be normalized so that each data indicator can be more equally weighted. Min-max normalization is utilized, where the minimum value in the range of values for each dataset is set at 0, the maximum value is set at 1, and every other value is transformed into a decimal between 0 and 1. For example, if the minimum value for unemployment was 10 and the maximum value was 30, a value of 20 would be transformed to 0.5 since it is halfway between 10 and 30.

Calculate cumulative index score

To combine all variables into a single cumulative index score, we average the normalized values across indicators.

Assign priority

Census block groups are sorted by their cumulative index score from highest to lowest. Census block groups that are in the top 25 percentile (i.e. have a cumulative index score in the 75 - 100th percentile) will be considered the prioritized communities.